A Risk and Comfort Optimizing Motion Planning Scheme for Merging Scenarios

摘要: Motion planning for merging scenarios accounting measurement and prediction uncertainties is a major challenge on the way to autonomous driving. Classical methods subdivide motion into behavior trajectory planning, thus narrowing down solution set. Hence, in complex scenarios, no suitable might be found. In this work, we present scheme that solves together by exploring all possible decision options. A safety strategy implemented risk of violating constraint minimized as well jerk feature comfort optimal trajectory. To mitigate injection noise actual trajectory, new analytical generation method derived its optimality proven. The capability evaluated through Monte-Carlo simulation. Furthermore, calculation time showing real-time our approach.